Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
FRONT PAGE
1 / 30
CONFERENCE ABSTRACTS
2018 2nd International Conference on Algorithms, Computing
and Systems (ICACS 2018)
2018 年第二届算法、计算与系统国际会议
Workshop
2018 2nd International Conference on Communications and
Future Internet (ICCFI 2018)
2018 年第二届通信与未来互联网国际会议
July 27-29, 2018 Beijing, China
2018 年 7 月 27-29 日 北京
Venue: BUPT Hotel Add: Beijing University of Posts and Telecommunications, 10 Xitucheng Rd, Haidian, Beijing
地址:北京海淀区西土城路 10 号
Published by Sponsored by
CONTENTS
2 / 30
Welcome Letter····························································································3
Notes and Tips·································································································4
General Agenda at a Glance···············································································5
Introduction of Keynote Speakers & Plenary Speaker·········································7
Speeches & Parallel Sessions··········································································11
Poster Sessions·······························································································28
WELCOME
3 / 30
Dear professors and distinguished delegates,
Welcome to 2018 2nd International Conference on Algorithms, Computing and Systems and
2018 2nd International Conference on Communications and Future Internet in Beijing, China!
We wish to express our sincere appreciation to all individuals and organizations who have
contributed to the conference. Special thanks are extended to our colleagues in the technical
program committee for their thorough review of all the submissions, which is vital to the success
of the conference, and also to the members in the organizing committee who had dedicated
their time and efforts in planning, promoting, organizing and helping the conference. Our special
thanks also go to the invited speakers as well as all the authors for contributing their latest
research to the conference.
This conference program is highlighted by four keynote speakers: Prof. Tok Wang Ling, National
University of Singapore, Singapore; Prof. Chin-Chen Chang, Feng Chia University, Taiwan; Prof.
Sheng-Uei Guan, Xi’an Jiaotong-Liverpool University, China; Prof. Jun Xu, Chinese Academy of
Sciences, China.
Oral presentations are divided into four parallel sessions. One best presentation will be selected
from each session, evaluated for: Originality, Applicability, Technical Merit, Visual Aids, and
English Delivery. We wish you all the best of luck with your presentations!
We believe that by this excellent conference, you can get more opportunity for further
communication with researchers and practitioners with the common interest in Algorithms,
Computing and Systems, Communications and Future Internet.
We wish you a pleasant and memorable experience in this conference as well as in Beijing.
Yours sincerely,
Conference Organizing Committee
Beijing, China
NOTES & TIPS
4 / 30
Notes:
You are welcome to register at any working time during the conference.
Please kindly keep your Paper ID in mind so that the staff can quickly locate your
registration information onsite.
Certificate of Listener can be collected in front of the registration counter. Certificate of
Presentation will be awarded after your presentation by the session chair.
One Best Presentation will be selected from each parallel session and the author of best
presentation will be announced and awarded when the session is over.
Your punctual arrival and active involvement in each session will be highly appreciated.
Please kindly make your own arrangements for accommodations.
Please keep all your belongings (laptop and camera etc.) with you in the public places, buses,
metro.
Warm Tips for Oral Presentation:
Get your presentation PPT or PDF files prepared.
Regular oral presentation: 15 minutes (including Q&A).
Laptop (with MS-Office & Adobe Reader), projector & screen, laser sticks will be provided by
the conference organizer.
AGENDA
5 / 30
< July 27, 2018>
Lobby
10:00-17:00 Registration & Materials Collection
Morning < July 28, 2018>
Function Hall (2
nd floor)
09:00-09:10 Opening Remarks
Prof. Sheng-Uei Guan, Xi’an Jiaotong-Liverpool University, China
09:10-10:10 Keynote Speech I
Prof. Tok Wang Ling
National University of Singapore, Singapore
Speech Title: Conceptual Modeling Views on SQL and NoSQL
10:10-10:30 Coffee Break & Group Photo
10:30-11:10 Keynote Speech II
Prof. Chin-Chen Chang
Feng Chia University, Taiwan
Speech Title: Applying De-clustering Concept to Information
Hiding
11:10-11:50 Keynote Speech III
Prof. Sheng-Uei Guan
Xi’an Jiaotong-Liverpool University, China
Speech Title: Opportunities and Challenges in Information
Communications Technology
11:50-12:30 Plenary Speech
Prof. Jun Xu
Chinese Academy of Sciences, China
Speech Title: Reinforcement Learning to Rank for Search Result
Diversification
Lunch Time <12:30-14:00> Location: Restaurant (2nd floor)
Note: lunch coupon is needed for entering the restaurant.
AGENDA
6 / 30
Afternoon< July 28, 2018>
14:00-15:45
Session I-Image and Signal Processing-7 presentations
Function Hall
1 (2nd
floor) AC011, AC024, AC028, FP012, FP022, AC002, AC206
14:00-16:00
Session II-Algorithm Design and System Modeling-8 presentations
Function Hall
2 (2nd
floor) AC008, AC023, AC026, AC036, AC039, AC042, AC202, AC031
Coffee Break <16:00---16:15>
16:15-18:45
Session III-Communication Engineering and Information Network-10 presentations
Function Hall
1 (2nd
floor) FP004, FP005, FP009, FP013, FP025, FP029, FP028, AC027, AC043, FP026
16:15-18:30
Session IV-Computer Science and Engineering-9 presentations
Function Hall
2 (2nd
floor) FP003, AC010, AC019, AC022, AC037, AC051, AC207, AC046, AC049
Poster Session AC006, AC016, AC205, AC050, AC204, FP019
Dinner <18:45-20:30> Location: Restaurant (2nd floor)
Note: dinner coupon is needed for entering the restaurant.
< July 29, 2018>
Beijing
08:00-17:00 Social Networking Event
The Event is optional and will charge an additional 80 USD/500 RMB. Participants should apply for it in advance
before July 10.
The Routes are:
Tian’ an men Square, Forbidden City, Summer Place and Wangfujing.
It includes transport to Tian’ an men square, the first tickets of the Palace Museum and the Summer Place, and
lunch.
KEYNOTE
7 / 30
Prof. Tok Wang Ling
IEEE Senior Life Member
National University of Singapore, Singapore
Dr. Tok Wang LING is a professor in Computer Science Department at the National University of
Singapore. He was Head of IT Division, Deputy Head of the Department of Information Systems
and Computer Science, and Vice Dean of the School of Computing of the University.
He received his PhD and M.Math, both in computer science, from University of Waterloo (Canada)
and B.Sc. (1st class Hons) in Mathematics from Nanyang University (Singapore).
His current research interests include Database Modeling, Entity-Relationship Approach,
Object-Oriented Data Model, Normalization Theory, Semi-Structured Data Model, XML Twig
Pattern Query Processing, XML and Relational Database Keyword Query Processing, Temporal
Database keyword Query Processing.
He serves/served on the steering committees of 5 international conferences, was the SC Chair of
both ER and DASFAA, and currently is the SC Chair of BigComp. He served as Conference Co-chair
of 12 international conferences, including ER 2004, DASFAA 2005, SIGMOD 2007, VLDB 2010,
BigComp 2015, and ER 2018. He served as Program Committee Co-chair of 6 international
conferences, including DASFAA 1995, ER 1998, ER 2003, and ER 2011.
He received the ACM Recognition of Service Award in 2007, the DASFAA Outstanding
Contributions Award in 2010, and the Peter P. Chen Award in 2011.
KEYNOTE
8 / 30
Prof. Chin-Chen Chang
IEEE and IET Fellow
Feng Chia University, Taiwan
Professor Chin-Chen Chang obtained his Ph.D. degree in computer engineering from National
Chiao Tung University. His first degree is Bachelor of Science in Applied Mathematics and master
degree is Master of Science in computer and decision sciences. Both were awarded in National
TsingHua University. Dr. Chang served in National Chung Cheng University from 1989 to 2005. His
current title is Chair Professor in Department of Information Engineering and Computer Science,
Feng Chia University, from Feb. 2005. Prior to joining Feng Chia University, Professor Chang was
an associate professor in Chiao Tung University, professor in National Chung Hsing University,
chair professor in National Chung Cheng University. He had also been Visiting Researcher and
Visiting Scientist to Tokyo University and Kyoto University, Japan. During his service in Chung
Cheng, Professor Chang served as Chairman of the Institute of Computer Science and Information
Engineering, Dean of College of Engineering, Provost and then Acting President of Chung Cheng
University and Director of Advisory Office in Ministry of Education, Taiwan.
Professor Chang's specialties include, but not limited to, data engineering, database systems,
computer cryptography and information security. A researcher of acclaimed and distinguished
services and contributions to his country and advancing human knowledge in the field of
information science, Professor Chang has won many research awards and honorary positions by
and in prestigious organizations both nationally and internationally. He is currently a Fellow of
IEEE and a Fellow of IEE, UK. And since his early years of career development, he consecutively
won Institute of Information & Computing Machinery Medal of Honor, Outstanding Youth Award
of Taiwan, Outstanding Talent in Information Sciences of Taiwan, AceR Dragon Award of the Ten
Most Outstanding Talents, Outstanding Scholar Award of Taiwan, Outstanding Engineering
Professor Award of Taiwan, Chung-Shan Academic Publication Awards, Distinguished Research
Awards of National Science Council of Taiwan, Outstanding Scholarly Contribution Award of the
International Institute for Advanced Studies in Systems Research and Cybernetics, Top Fifteen
Scholars in Systems and Software Engineering of the Journal of Systems and Software, Top Cited
Paper Award of Pattern Recognition Letters, and so on. On numerous occasions, he was invited to
serve as Visiting Professor, Chair Professor, Honorary Professor, Honorary Director, Honorary
Chairman, Distinguished Alumnus, Distinguished Researcher, Research Fellow by universities and
research institutes. He also published over hundreds papers in Information Sciences. In the
meantime, he participates actively in international academic organizations and performs advisory
work to government agencies and academic organizations.
KEYNOTE
9 / 30
Prof. Sheng-Uei Guan
Xi’an Jiaotong-Liverpool University, China
Steven Guan received his BSc. from Tsinghua University (1979) and M.Sc. (1987) & Ph.D. (1989)
from the University of North Carolina at Chapel Hill. He is currently a Professor and the Director
for Research Institute of Big Data Analytics at Xi'an Jiaotong-Liverpool University (XJTLU). He
served the head of department position at XJTLU for 4.5 years, creating the department from
scratch and now in shape. Before joining XJTLU, he was a tenured professor and chair in
intelligent systems at Brunel University, UK.
Prof. Guan has worked in a prestigious R&D organization for several years, serving as a design
engineer, project leader, and department manager. After leaving the industry, he joined the
academia for three and half years. He served as deputy director for the Computing Center and
the chairman for the Department of Information & Communication Technology. Later he joined
the Electrical & Computer Engineering Department at National University of Singapore as an
associate professor for 8 years.
Prof. Guan’s research interests include: machine learning, computational intelligence, big data
analytics, mobile commerce, modeling, networking, personalization, security, coding theory, and
pseudorandom number generation. He has published extensively in these areas, with 130+
journal papers and 180+ book chapters or conference papers. He has chaired, delivered keynote
speech for 80+ international conferences and served in 180+ international conference
committees and 20+ editorial boards. There are quite a few inventions from Prof. Guan including
Generalized Minimum Distance Decoding for Majority Logic Decodable Codes, Prioritized Petri
Nets, Self-Modifiable Color Petri Nets, Dynamic Petri Net Model for Iterative and Interactive
Distributed Multimedia Presentation, Incremental Feature Learning, Ordered Incremental
Input/Output Feature Learning, Input/Output Space Partitioning for Machine Learning, Recursive
Supervised Learning, Reduced Pattern Training using Pattern Distributor, Contribution Based
Feature Selection, Incremental Genetic Algorithms, Incremental Multi-Objective Genetic
Algorithms, Decremental Multi-objective Optimization, Multi-objective Optimization with
Objective Replacement, Incremental Hyperplane Partitioning for Classification, Incremental
Hyper-sphere Partitioning for Classification, Controllable Cellular Automata for Pseudorandom
Number Generation, Self Programmable Cellular Automata, Configurable Cellular Automata,
Layered Cellular Automata, Transformation Sequencing of Cellular Automata for Pseudorandom
Number Generation, Open Communication with Self-Modifying Protocols, etc.
PLENARY
10 / 30
Prof. Jun Xu
Chinese Academy of Sciences, China
Prof. Jun Xu received his B.E. and Ph.D. in Computer Science from Nankai University, in 2001 and
2006, respectively. He worked as an associate researcher, researcher, and senior researcher at
Microsoft Research Asia and Huawei Noah's Ark Lab. In 2014, he joined Institute of Computing
Technology, Chinese Academy of Sciences. His research interests are in information retrieval,
machine learning, and big data analysis. I have worked on (i) learning to rank for information
retrieval; (ii) large scale topic modeling; and (iii) semantic matching in search.
ABSTRACT
11 / 30
July 28, 2018
Opening & Speeches
Time: 09:00-12:30
Function Hall (2nd floor)
09:00-09:10 Opening Remarks
Prof. Sheng-Uei Guan, Xi’an Jiaotong-Liverpool University, China
09:10-10:10
Conceptual Modeling Views on SQL and NoSQL
Prof. Tok Wang Ling
National University of Singapore, Singapore
Abstract: The three basic concepts in the Entity Relationship Model are: object class,
relationship type, and attribute of object class and relationship type. They are termed
ORA-semantics. Without knowing the ORA-semantics in the databases, the quality of
some databases are low. The Relational Model does not capture ORA-semantics.
We first present some restrictions in the relational model such as limitations of
normalization theory and limitations of Relational Model using the Universal Relation
Assumption. We then discuss the requirements for traditional database applications in
RDBMS using SQL and the performance issues, review some of the existing methods
which can be used to improve the performances of certain database applications, such
as materialized view design, horizontal and vertical partitioning of data, the concepts of
strong FD/MVD and weak FD in physical database design, etc.
We further present some rarely-mentioned but very important issues in data and
schema integration, such as entity resolution vs relationship resolution, primary key vs
object identifier (OID), local OID vs global OID, local FD/MVD vs global FD/MVD,
semantic dependency vs FD/MVD, etc. All these concepts are related to ORA-semantics
and they have significant impact on the quality of the integrated database and schema.
We briefly present the basic data models of the 4 major categories of NoSQL databases,
i.e. key-value store, wide-column store, document store, and graph database. We then
give a conceptual modeling view on SQL vs NoSQL using a set of characteristics and
requirements of database applications. We finally present a set of criteria to aid
decision-making regarding when to use SQL or NoSQL for database applications.
Coffee break & group photo
10:10---10:30
10:30-11:10
Applying De-clustering Concept to Information Hiding
Prof. Chin-Chen Chang
Feng Chia University, Taiwan
ABSTRACT
12 / 30
Abstract: Reversible steganography allows an original image to be completely restored
after the extraction of hidden data embedded in a cover image. In this talk, I will talk
about a reversible scheme based on declustering strategy for VQ compressed images.
The declustering can be regarded as a preprocessing step to make the proposed
steganographic method more efficient. Our experimental results show that the time
required for the embedding process in the proposed method is few. In addition, the
reversible steganography allows an original image to be completely restored after the
extraction of hidden data embedded in a cover image. In this talk, l will introduce a
reversible scheme for VQ-compressed images that is based on a declustering strategy
and takes advantage of the local spatial characteristics of the image. The main
advantages of our method are ease of implementation, low computational demands,
and no requirement for auxiliary data.
11:10-11:50
Opportunities and Challenges in Information Communications Technology
Prof. Sheng-Uei Guan
Xi’an Jiaotong-Liverpool University, China
Abstract: This talk introduces the overall trends of Information Communications
Technology (ICT) and presents an overview for opportunities and challenges in ICT.
Critical issues, research problems and developments of ICT in various areas are
addressed, such as green computing, Internet computing, mobile computing, and
intelligent computing. Opportunities and challenges in relevant areas are also covered,
for example, Internet of Things, cloud computing, big data analytics. Critical
development of ICT in various aspects are proposed thereafter. Finally, the challenges
faced by the higher education sector are also discussed.
11:50-12:30
Reinforcement Learning to Rank for Search Result Diversification
Prof. Jun Xu
Chinese Academy of Sciences, China
Abstract: The goal of search result diversification is to construct a document ranking
for satisfying as many different query subtopics as possible. Typically, the diverse
ranking process can be formalized as greedy sequential document selection. At each
position, the document that can provide the largest amount of additional information
to the users is selected. Since the utility of a document depends on its preceding
documents in search result diversification, constructing an optimal document ranking is
NP-hard. The traditional greedy document selection usually leads to suboptimal
solutions. In the talk, I will show that the problem can be alleviated with a Monte Carlo
tree search (MCTS) enhanced Markov decision process (MDP) model. Specifically, the
sequential document selection process is fit into an MDP. At each time step the greedy
action is further improved through the exploratory tree search by MCTS.
Reinforcement learning algorithm was developed to learn the model parameters.
ABSTRACT
13 / 30
Empirical evaluation clearly indicated the effectiveness of the approach. The MCTS
enhanced MDP can also be applied to variant applications, including sequence tagging,
text matching etc.
Lunch Time <12:30-14:00> Location: Restaurant (2nd floor)
Note: lunch coupon is needed for entering the restaurant.
Session I: Image and Signal Processing
Time: 14:00-15:45
Function Hall 1 (2nd floor)
Chair: Prof. Kwan Hyeong Lee, University of Daejin, South Korea
AC011
14:00-14:15
Optimized Software-Based Hardening Strategies for Matrix Multiplication and Fast
Fourier Transform
Zhijian Hui, Yangsheng Wang, Honghui Tang, Tao Qin and Haibin Wang
Hohai University, China
Abstract—Nowadays, Graphics Processing Unit (GPU) has shown great potential in
High-Performance Computing applications for its parallel computing structures, which
can greatly accelerate the computing process. However, GPU reliability is critical in
some applications like satellite or auto-driving. Thus, many researches have been
carried out to improve its reliability using both hardware and software schemes. In this
paper, we mainly focus on designing schemes to protect devices from soft errors. We
analyze the performance of available algorithm-based fault tolerance (ABFT) schemes
for two commonly used mathematical operations: Matrix Multiplication and Fast
Fourier Transform (FFT). We developed optimized schemes to reduce the overhead of
the available ones. The proposed schemes are based on two insights. First, in matrix
multiplication, the scheme may classify some correct results as potential errors when
there are multiple errors, introducing unnecessary overhead. Second, in FFTs, the
overhead of the ABFT scheme depends on its size, so we can use butterfly module to
detect errors and correct them. Finally, we use fault-injection simulation to evaluate the
proposed schemes and they are proved to perform quite well in any error distributions.
AC024
14:15-14:30
Human tracking using TLD with Automatic Initiation
Jiawei Shi, Xianmei Wang and Huer Xiao
University of Science and Technology Beijing, China
Abstract—To realize human tracking, a framework by TLD tracking algorithm and
ABSTRACT
14 / 30
dynamic average background modeling is shown in this paper. First, totally
automatically human initiation module is given by background modeling algorithm and
classification, which output candidate and confirmed human patches. Then TLD
framework is employed to track each human object until it disappear. Both of the
output of tracking and initializer are used together to decide how to further update the
tracking list. Experiments results on PETS show the performance of our solution.
AC028
14:30-14:45
Multi-Steps Weighted ARMA Identification Algorithm for the Multi-sensors System with
Unknown Parameter
Li Heng, Hang Yan-Yan, Zhang Chen-Chen, Yin Jia-Lai, Guan Biao And Sun Hui-Fen
Fuyang Normal University, China
Abstract—For the multi-focus image fusion, determining which pixels or regions are
located in focus is one of the difficulties. In the field of image fusion, wavelet analysis is
becoming a hot research topic. Fusion rules of low frequency sub-band and high
frequency sub-bandare becoming the focus and difficulty. At present, the traditional
fusion rules are not perfect. With no thought for a relationship between low frequency
and high frequency sub-band coefficients, it is likely to choose the wrong coefficients.
Aiming at the existing problems, after comparing several traditional fusion rules, a new
fusion rule is proposed. Experimental results indicate that the algorithm is of low
complexity, and the quality of the fused image is better.
FP012
14:45-15:00
Data Visualization of Abaca Production in Catanduanes
Belen M. Tapado andThelma D. Palaoag
Catanduanes State University, Philippines
Abstract—This research draws inspiration from the desire to help sustain and boost the
production of abaca in Catanduanes being the topmost abaca producing province in the
Philippines through an information campaign in a website with the application of data
visualization tools and techniques. Result of this research will draw attention about the
usefulness and benefits that could be derived through the production of abaca in
uplifting the socio-economic status and standard of living of farmers. Likewise, this
research would also support the rural development program of the government on
embracing climate-resilient ecological farming practices to adjust to climate change.
Sustainability of the abaca production in the province means ensuring vibrancy of
economy especially for the abaca farmers. Demographic data about abaca production
were obtained from the abaca concerned government offices in Catanduanes.
Documentary analysis, Interview, Website development and evaluation were the
methodologies employed in this investigation. Likewise, a survey questionnaire was
utilized to gather data about the usability of the website from the persons from
government agencies that are concerned on abaca and website development. Three (3)
criteria were used to evaluate the usability of the website – usefulness, ease of use and
satisfaction. Frequency count and weighted mean were the statistics used in analyzing
the responses and the respondents have generally evaluated the developed website as
ABSTRACT
15 / 30
“Strongly Agree”.
FP022
15:00-15:15
Speech Enhancement Using Successive State Estimation under Industrial Noise
Environment
Qinghe Wu, Haifeng Wu and Yu Zeng
Yunnan Minzu University, China
Abstract—Speech communication under an industrial noise environment can seriously
affect an operator's hearing and collaboration. For this environment, we use a
successive state estimation method to perform speech enhancement to eliminate the
effect of noise on speech. The successive state estimation uses Kalman filter algorithm.
Compared with traditional Kalman speech-enhancement methods, we use a simpler
difference form for a state equation to reduce the number of equation parameters. In
experiments, we adopt a noise signal in a public database and add it to an actual
speech signal as a noisy speech signal to be processed. Compared with Fast ICA
(Independent Component Analysis) algorithm, this speech enhancement algorithm has
better real-time performance.
AC002
15:15-15:30
Differences Analysis of HRV in Time and Meridians Based on Multi-lead ECG Signals,
Jihong Liu, Xinyu Xu, Guangwei Zhang, Simeng Wu,Mingzhen Liu and Yuning Zhou
Northeastern University, China
Abstract—In Traditional Chinese Medicine theory, meridian is a connection, regulation
and reaction system of human functional activities. The Qi-Blood Circulation of the
veins has its ups and downs, one day is divided into 12 two-hour periods, and one
two-hour period is distributed one meridian. HRV analysis index has a wide range of
applications, including in physiology, cardiology, sports, health, mortality prediction,
sleep and pain, and used in the study of medicine and exercise routine. Based on
multi-lead ECG signals, the differences of heart rate variability (HRV) indexes between
heart meridian of Hand-Shaoyin and pericardium meridian are analyzed in this paper. In
different meridians and time, the differences of SDNN, PNN50, rMSSD and other HRV
indexes are discussed. The experimental system is designed in MATLAB platform.
AC206
15:30-15:45
Text Prompted Speaker Verification Based on Phoneme Clustering with Earth Mover's
Distane and Cauchy-Schwarz Divergence
Zhuzi Chen, Yi Liu and Jia Liu
Tsinghua University, China
Abstract—For short duration text prompted speaker verification where the amount of
enrollment data is limited for each speaker model, it is hard to obtain a robust speaker
representation. In these situations of short utterance speaker verification
I-vector/GMM approaches work even worse than traditional GMM-MAP modeling
method. GMM/HMM framework content matching is the state-of-the-art paradigm for
short duration text-dependent speaker verification, in which models for individual
lexical such as words, syllables, or phonemes are established for the background and
ABSTRACT
16 / 30
speaker to make up mismatch. However, some of the phonemes do not occur in
enrollment but happen in the testing recordings, and most of the phonemes have
different preceding and succeeding phonemes, both of which leads to coarticulation
difference. These are called lexical and context mismatch. In this work, to overcome the
data sparceness caused lexical mismatch and context mismatch, phoneme states are
clustered applying Earth Mover’s Distance and Cauchy-Schwarz divergence as metrics.
Performance improved as EER lowered by 6.2%, minDCF08 lowered by 1.9% for Earth
Mover’s Distance metric, and EER lowered by 3.7%, minDCF08 rised 1.9% for
Cauchy-Schwarz divergence metric.
Session II: Algorithm Design and System Modeling
Time: 14:00-16:00
Function Hall 2 (2nd floor)
Chair: Prof. Tok Wang Ling, National University of Singapore, Singapore
AC008
14:00-14:15
An Extended Model for Simulation of Mexican Wave
Libi Fu, Shuchao Cao and Jie Fang
Fuzhou University, China
Abstract—“Mexican wave” is an interesting phenomenon in a stadium, and has
attracted much attention. Some researchers have established models to reproduce it,
and related mechanisms have been understood. However, the influence of some
factors such as density and individuals’ visual field on this process has not been well
investigated. Hence, this paper aims to simulate wave propagation by modifying a
model introduced by Farkas et al. Detailed rules are given. After involving the effect of
average density, it is observed that there will be a decrease in activation threshold if the
average density decreases. This result is reasonable, and can be analyzed by
calculations. Then, spectators’ visual field is considered in the model. The influence of
initial locations of triggering group members on activation thresholds is discussed.
Results demonstrate that it is easier to generate a wave at a relatively large average
density if the location is in front rows of a stadium. It is hoped that this study will be
helpful in a deep understand of the propagation mechanisms of Mexican waves and in
crowd management.
AC023
14:15-14:30
A Cooperative Artificial Bee Colony Algorithm and its application to Fashion Color
Forecast in Clothing
Li Zhao, Yang Lianhe, Zhang Shaoqiang and Zhang Shaoqiang
Tianjin Normal University, China
Abstract—This paper presents a hierarchical cooperative artificial bee colony algorithm
based on divide-and-conquer decomposition strategy (HCABC-D), for fashion color
ABSTRACT
17 / 30
forecast in clothing. In the proposed algorithm, classical artificial bee colony is extended
to cooperative and hierarchical structure. The top level is responsible for information
aggregation from lower level and information exchange based on crossover operator. In
the bottom level, each sub-population also adopts the canonical ABC algorithm to
search the part-dimensional landscape. Furthermore, HCABC-D and ABC are applied in
forecasting fashion color in clothing. The results show that HCABC-D provides extremely
competitive performance. The comparison between forecasting results and ones issued
by PANTONE Inc. demonstrates its performance superiority.
AC026
14:30-14:45
Airfoil Optimization based on Isogeometric DG
Kun Wang, Shengjiao Yu and Tiegang Liu
Beihang University, China
Abstract—In this paper, an adjoint-based airfoil optimization algorithm is developed
based on isogeometric discontinuous Galerkin (IDG) method for compressible Euler
equations. We first parameterize the airfoil by B-spline curve approximation with some
control points viewed as design variables, and build the B-spline representation of the
flow field with the curve to apply global refined IDG method for flow solution. With the
isogeometric nature, not only all the geometrical cells but also the numerical basis
functions can be analytically expressed by the design variables. Consequently, the
gradient involved in SQP optimization algorithm is totally estimated in an accurate
approach indicating that the numerical solutions and objective could be differentiable
with respect to those variables. The proposed algorithm is demonstrated on RAE2822
airfoil with inviscid transonic flow.
AC036
14:45-15:00
Equivalent State of Finite Automata and Its Judgment Theorem
Xingyan Yue, Peifeng Hao and Zhao Lin
Northeastern University, China
Abstract—Automata theory is an important tool in modern computer science, playing a
role in compiler and many other engineering fields. This paper first analyzes the
automata theory. For sake of enhancing the mathematical logicality of automata theory
and to carry out the further research on the equivalent state of finite automata, we
modify and optimize the basic concept of automata theory. Next, this paper defines the
string set of state, in order to formalize the concept of equivalent state. According to
these basic concepts, we advance several judgment theorems of equivalent state,
which can explain the effectiveness of the simplification algorithms of finite automata,
and complete their proofs. Finally, we also introduce a merging based simplification
algorithm for finite automata, which is the direct application of one of the judgment
theorems above.
AC039
15:00-15:15
On-line Optimization of Energy-saving Train Control using Bacteria Foraging Algorithm
Xiao Siyu, Liu Jiang and Cai Baigen
Beijing Jiaotong University, China
ABSTRACT
18 / 30
Abstract—Reducing the energy consumption has been a significant issue in urban
transit operations. Considering the prior constraint from the operational schedule, the
online optimization of the speed curve between successive stations is capable of
improving the energy saving capability and enhancing cost-efficiency of the whole rail
transit system. To fulfill the requirements of schedule adherence, an on-line
optimization method for reducing energy consumption the train traction control
operation is presented in this paper. In this method, the bacteria foraging algorithm is
adopted to achieve an optimal solution of an objective function concerning the energy
consumption and other related factors, with which the speed curve for the following
rail section can be updated and adjusted in-time to cope with the possible deviation
between time schedule and the practical operation. With the integration of the rail
dispatching and train operation control, the control strategy of the train can be
updated according to specifically modified trip plan, which means that the changed
schedule for the following rail section still can be fulfilled by re-calculating the
optimized speed curve of the train within the standing time in a station. The
optimization is achieved under a multi-objective optimization framework, where the
energy consumption and passenger comfort degree are concerned under
atime-domain constraint that might be adjusted based on the latest schedule. Results
from simulation demonstrate the effectiveness and feasibility of the proposed method,
which illustrate the great potentials in safer and greener urban transit systems in the
future.
AC042
15:15-15:30
An Adaptive Step Size Based Algorithm for Hybrid Model Reliability Analysis
Tao Qiu, Jianguo Zhang, Juan Wei and Lingfei You
Beihang University, China
Abstract—Focusing on the mixed conditions of both random variables and interval
variables, this paper constructs a hybrid reliability analysis algorithm based on adaptive
step size. First, the double-loop optimization model is decoupled into a sequence
reliability analysis model of probability analysis and interval analysis. In probability
analysis, we use the existing iterative information to determine the adaptive step
length parameters and find the most probability point (MPP). In interval analysis,
quadratic programming is used to find the optimum point in the interval variable space.
Finally, the results of two case studies show that the proposed algorithm is compared
with Monte Carlo Sampling (MCS). The relative error is within 5%, and the number of
iterations can reach convergence within 100 times. In particular, when the performance
function has a high degree of nonlinear, the algorithm also has high calculation
accuracy, efficiency and convergence stability.
AC202
15:30-15:45
Heuristic Based Scheduling for Tablets Film Coating Process
Chutamas Pontrakul and Pisit Jarumaneeroj
Chulalongkorn University, Thailand
Abstract—In the pharmaceutical industry, Tablets Film Coating Planning concerns a
ABSTRACT
19 / 30
finding of coating sequences on parallel machines such that both the completion time
and number of tardy jobs are minimized. This problem is made more difficult when a
wide variety of drugs is produced as changing from one to another drug requires
considerably long period of cleaning and setup times – these times also vary depending
on the previously coated drug. We show that this problem could be transformed into
the Vehicle Routing Problem with Time Window (VRPTW), where a variable arc
exchange heuristic − whose concept is based on the Variable Neighborhood Search – is
devised to solve such a problem. Our proposed heuristic is a two-phase one, where the
initial solutions are constructed from easy-to-implement dispatching heuristics, i.e.
Earliest Due Date (EDD) and Longest Processing Time (LPT). The initial solution is then
iteratively improved by a series of local-search operators, including 2OPT, RELOCATION,
and SWAP. The results from our proposed heuristic is comparatively good, when
compared to those of the optimization model in terms of solution quality as the gap is
less than 5% for instances of 20 orders; but, it requires much less computation time,
which is crucial from a practical point of view.
AC031
15:45-16:00
A Dual Alternating Direction Method of Multipliers for the Constrained Lasso Problems
Wang Qingsong
Southwest Jiaotong University, China
Abstract—In this paper, we developed a dual alternating direction method of
multipliers (ADMM) for the constrained lasso problem, which is the standard lasso
problem with linear inequality constraints. Under mild conditions, we present the
global convergence and local linear convergence rate for the algorithm. Numerical
experiments demonstrate the efficiency and robustness of the dual ADMM.
Coffee Break <16:00---16:15>
Session III: Communication Engineering and Information Network
Time: 16:15-18:45
Function Hall 1 (2nd floor)
Chair: Assoc. Prof. Mischelle A. Esguerra, Lyceum of the Philippines University-Batangas, Philippines
FP004
16:15-16:30
Virtual Assistant for IoT process management, using a middleware
David Chilcañán, Patricio Navas and Milton Escobar
Universidad de las Fuerzas Armadas ESPE, Ecuador
Abstract—The implementation of virtual assistants, the automation of homes, smart
cities, management of remote sensors, among other technologies related to the
ABSTRACT
20 / 30
internet of things (IoT). They allow to improve the control, monitoring and
management of your processes. The present research work employs a Message
Oriented Middleware (MOM) for interconnecting devices allowing to monitor the
domestic energy consumption of a household (KWH, fees to be paid by day, week,
month) in real time, using a wizard virtual (chatbot) which facilitates the Administration
and control of the connected electrical devices. Through a combination of ubiquitous
distributed sensing units. Adding data, reasoning and awareness of the context.
FP005
16:30-16:45
On the spectral efficiency of full-duplex massive MIMO system with low-resolution
ADCs/DACs
Mo Xiaohao, Xu Kui and Ma Wenfeng
The Army Engineering University of PLA, China
Abstract—This paper focuses on a full-duplex (FD) massive multi-input multi-output
(MIMO) system with low-resolution analog-to-digital (ADC) and digital-to-analog (DAC),
where base station (BS) is equipped with large-scale uniform linear antenna array
(ULA). Our goal is to evaluate the impact of using low-resolution ADCs/DACs on system
performance. We analyze the asymptotic uplink and downlink spectral efficiency (SE) of
the receive signal with zero-forcing (ZF) and maximum ratio combination/transmission
(MRC/MRT) methods. In simulation part, we know that the spectral efficiency rises with
the number of antennas at BS and resolution of ADC and DAC. Most importantly, we
found the performance of using 5-bit ADCs and DACs is satisfying compared with
perfect ADCs and DACs.
FP009
16:45-17:00
Adaptive Modulations for OFDM-MIMO systems
Guowei Lei and Xuefang Xiao
Jimei University, China
Abstract—Orthogonal frequency division multiplexing (OFDM) is regarded as a popular
technique. Moreover, OFDM combined with multiple-input multiple-output systems
(MIMO) has drawn many interests. However, OFDM-MIMO systems with various
modulations have not been investigated extensively to date. This letter is to draw a
comparison among the linear and non-linear modulations. Especially, the random
modulation index is introduced into the OFDM-MIMO system with CPM. Meanwhile
the OFDM-MIMO system with antenna selection and adaptive modulation is proposed.
The performances for the proposed systems are evaluated via simulation.
FP013
17:00-17:15
GIS BASED MANGROVE LOCATION MAPPING AND INFORMATION SYSTEM IN
CATANDUANES
Gemma Acedo and Thelma Palaoag
Catanduanes State University, Philippines
Abstract—Catanduanes has a vast home grown mangrove species. Waterfront and
mangrove in Catanduanes Island in Luzon, Philippines (Lat 13.67 °, Long 124.12 °) is
conceivable along the eastern, western and southern coasts. There are several
ABSTRACT
21 / 30
shorelines and ocean shores that possess marine greenery (marine full scale green
growth) and fauna. This paper displays a geographic data framework (GIS) - based
multi-criteria basic guidance approach for mangrove species. This assesses the relative
needs of mapping and identifying the different kinds of mangrove in light of an
arrangement of inclinations, criteria and markers for the province. It determines
whether the government should enhance their effort to improve the ecological habitat
in mangrove reforestation that shall have a great impact on the province with regards
to environment, climate change on the marine nature in a natural way of life. This
Creation of geographic information system in the province will help the environment to
preserve the habitat for different kinds of marine life and develop mangrove protected
areas.
FP025
17:15-17:30
Flow Control and Resource Allocation in Cognitive Radio Networks
Jain-Shing Liu and Wan-Ling Chang
Providence University, Taiwan
Abstract—In this work, we aim to optimize the system throughput of second user (SU),
while satisfying the collision constraint of primal user (PU), throughput constraint of
second user (SU), and scheduling constraint in interweave cognitive radio networks
(CRNs) under time-varying channel and traffic conditions. Our solution to the stochastic
optimization problem is realized by a flow control and resource allocation algorithm
based on the Lyapunov drift-plus-penalty framework that can resolve the time-varying
conditions without a-prior knowledge to obtain the long-term stability of the virtual
queues involved. In particular, the scheduling algorithm embedded possesses
polynomial time efficiency, which is highly desired for a practical system. Finally, our
simulation studies show the performance tradeoff between throughput and delay to
be [O(1/V),O(V)] with a system parameter V while satisfying all the requirements and
constraints under consideration.
FP029
17:30-17:45
A New Pilot-based Block Low-rank Channel Estimation Algorithm for Massive MIMO
Ting Fang and Guigen Zeng
Nanjing University of Posts and Telecommunications, China
Abstract—A new block low-rank channel estimation algorithm is put forward for the
problem that the complexity of the channel estimator matrix inverse computation for
the traditional linear minimum mean square error (LMMSE) channel estimator in the
massive MIMO system increases as the number of antenna increases. The block
partitioning algorithm extracting the key information of the channel autocorrelation
matrix taking the relevant bandwidth as a criterion and the low-rank estimation
method by use of the frequency domain (and/or time-domain) correlation and singular
value decomposition of the channel are employed. The channel autocorrelation matrix
is divided into several blocks according to the block size calculated by the channel
dependent bandwidth. In the inverse process, only the low frequency diagonal
sub-arrays representing the main information amount of the channel are applied, while
ABSTRACT
22 / 30
other sub-arrays that represent the channel high-frequency information have been
ignored. Rank estimation, namely the determination of signal subspace dimension,
should be considered with the compromise between the computational complexity
and the estimation error. Under the slow fading channel with selective frequency, this
algorithm compares the performance and computational complexity with the LMMSE
and the block LMMSE estimation algorithm. The result shows that the algorithm
reduces the computational complexity without loss of its performance.
FP028
17:45-18:00
20Gbps MDM-based Optical Multimode System for Short-haul Communication
Sushank Chaudhary, Xuan Tang, Bangjiang Lin and Xian Wei
Chinese Academy of Sciences, China
Abstract—Mode division multiplexing or MDM is a significant technology that is used
for increasing transmission capacity of multimode communication systems. MDM
allows the transmission of many channels over multimode link by using the
multiplexing of different modes with a single wavelength. This reduces the required
bandwidth for optical communication systems. Multimode fiber (MMF) and Few mode
Fiber (FMF) are the possible optical fibers used to carry the different modes from one
place to another place. The current study aims at designing a high-speed multimode
transmission system by using MDM of two modes – Laguarre Gaussian (LG) 00 and 01
modes – for transporting two independent channels each with 10 Gbps data over 6800
multimode link. Furthermore, comparative analyses of performance of few mode fiber
(FMF) and multimode fiber (MMF) are carried out in terms of Bit error rate (BER),
Quality factor and wye diagrams. Moreover, modal decomposition of LG 00 and 01
modes into linearly polarized (LP) modes is also reported.
AC027
18:00-18:15
A New Fusion Rule of Multi-focus Image Fusion Based on Wavelet Transform
Guan Biao, Zhang Chen-Chen and LI Heng
Fuyang Normal University, China
Abstract—For the multi-sensors time-invariant system with unknown parameters, in
order to improve the accuracy of the identification, based on the ARMA model, a kind
of multi-steps identification algorithm is presented. Step a: we use recursive extended
least squares method to get the local estimates of the unknown parameters, we use
arithmetic mean to get the first fusion estimates of the parameters. Step b: we use
correlated function method to get the estimates of observation noises of every single
sensor. Step c is the key step of our algorithm which differs from the conventional
2-steps algorithm. Step c: we take the related information of the estimates of every
single sensor as the weight, we take the weighted fusion of the local estimates as the
final estimates of the parameters. Compared to the real values, the final estimates are
more accurate than the first estimates. We use Matlab to simulate an typical example,
the simulation results show that the final estimates have better convergence than the
first estimates, which could show the accuracy advantage of the multi-steps
identification algorithm presented in this paper.
ABSTRACT
23 / 30
AC043
18:15-18:30
Android-Based Mobile Application for Swine Traceability and Management System
using Near Field Communication (NFC)
Mischelle A. Esguerra and Aubrey Rose M. Cordero
Lyceum of the Philippines University-Batangas, Philippines
Abstract—The widespread use of Information and Communications Technology (ICT) in
day-to-day life of the people has become common. In recent years, the rapid
emergence of Internet and mobile-based technologies were observed in urban and
rural communities, providing an easy and rapid delivery of information. For this
purpose, the authors developed an Android-Based Mobile Application for Swine
Traceability and Management System for a training center on pig husbandry using Near
Field Communication (NFC) technology.
This mobile application is of great help to the training centers of pig husbandry in
keeping tracks of the movement of swine from their farm to the slaughterhouse. By
using the mobile application, the training center had an easy way of identifying and
recording the pig’s information and maintains an organized record management
system. The use of Near Field Communication Technology provides the client an
innovative way of swine management, and is way cheaper than RFID.
This mobile application was developed using Android platform and runs on
NFC-compatible handsets with Jellybean 4.2 versions or higher. The application include
several features like offline work, personalization options and touch support. To ensure
the security of the data, options to backup, restore and export data were included in
the settings menu of the application.
FP026
18:30-18:45
Performance of Flight Target Position Estimation in Noise and Interference
Kwan Hyeong Lee
University of Daejin, South Korea
Abstract—This paper studied for direction of arrival estimation to use MUSIC algorithm
and optimum adaptive beamforming algorithms in a coherent interference
environment. Proposed method of this paper not increase the computational
complexity of MUSIC algorithm and adaptive beamforming to compare classical
subspace techniques. The proposed method is combined the recently updated weight
algorithm with LCMV beamforming algorithm in adaptive array system for direction of
arrival estimation of desired signal. It has great potential in case of mobile wireless
system where coherent and co-channel multipath interference is a major problem. The
proposed method can be used in conjunction with MUSIC algorithm to provide an
initialization for the weight value method to get a more accurate direction of arrival
estimation. Through simulation, we compare the LCMV method with classical direction
of arrival method based on MUSIC algorithm. We show that the propose method has
achieve good resolution performance better than classical direction of arrival
estimation algorithm.
ABSTRACT
24 / 30
Session IV: Computer Science and Engineering
Time: 16:15-18:30
Function Hall 2 (2nd floor)
Chair: Prof. Wei-Mei Chen, National Taiwan University of Science and Technology, Taiwan
FP003
16:15-16:30
Research on the Security and Guarantee of Computer Communication Network
Li HongXia and Chen Lei
Xi'an University of Science and Technology, China
Abstract—With the continuous development of science and technology, socialist
modernization has been gradually implemented. The extensive coverage area of the
communication network has made the computer communication network technology
in our country well developed and applied. Therefore, computer communication has
become a part of people's daily life. Under the background of the ever-increasing scale
of communication networks, communication network technology has brought
convenience and fun to people's work and life. At the same time, computer
communication network security issues have quietly come to people's side. Once there
is a problem of communication network security, it will lead to equipment failure, and
it will make the communication network system paralyzed. This will cause losses to
users and even the entire society. Therefore, it is very important to strengthen the
security of computer communication networks. This article mainly introduces the key
technologies of computer communication network security, analyzes the development
status and problems of communication network security, and discusses the effective
ways to ensure the security of communication networks from the three levels of the
country, the enterprise, and the user, hoping to strengthen the security of
communication networks.
AC010
16:30-16:45
Parallel Implementation of Simultaneous Perturbation Stochastic Approximation with
Adaptive Step Sizes
Yue Fan and Tiegang Liu
Beihang University, China
Abstract—The simultaneous perturbation stochastic approximation (SPSA) is an
iterative gradient-free optimization algorithm. This algorithm approximates the optimal
solution by estimating the gradient of the loss function and uses only 2 values of the
loss function to estimate the gradient which is independent of the dimension of the
problem. But for the high dimensional problem with large-scale, the convergence of
classical SPSA algorithm is slow, even can not converge to the global minimum value. In
order to accelerate the convergence of SPSA algorithm and improve the precision of the
optimum solution, this paper proposes an improvement of SPSA algorithm based on
the adaptive step sizes, which can adjust the step size of each iteration according to the
gradient direction. The modified algorithm is implemented by Message Passing
ABSTRACT
25 / 30
Interface (MPI) parallel in each iteration to widen the search scope for the better
descending direction in order to accelerate the convergence. Numerical experiments of
several classical nonlinear optimization problems shows that the modified SPSA
algorithm can accelerate the convergence of nonlinear optimization problems and find
the global extremum point of high-dimension for some optimization problems which
the SPSA algorithm can not find. It proves that the modified algorithm with adaptive
step sizes outperforms the classical SPSA algorithm.
AC019
16:45-17:00
Text Sentiment Polarity Classification Method Based on the Word Embedding
Sun Xiaojie, Du Menghao, Shi Hua and Huang Wenming
Guilin University of Electronic Technology, China
Abstract—Most of the machine learning algorithms for text sentiment analysis use the
word embedding obtained by word2vec training as their inputs. However, the word
embedding of word2vec training contains only semantic information. An algorithm for
text sentiment analysis is proposed solve the problem of text containing semantics,
syntax, sentiment and other information. It begins with the learning of original
text-multi word embedding in the semantic, syntactic, and sentiment information,
followed by proceeding the word embedding fusion. The improved convolution neural
network is applied for sentiment analysis. Thus, it solves the problem that the word
embedding contains monotonous text information. K-means text clustering is applied
by dividing similar text into the same cluster, thus improving the classification accuracy.
The application of the Principal Component Analysis (PCA) dimensionality not only
extracts the principal component information, but also solves the problem of
redundancy embedding and improves the computational performance of classification
model. The experiment results show that the presented method has a significant
improvement in the accuracy, recall rate and F value of the sentiment polarity analysis
of the critical text in comparison with other fusion algorithms.
AC022
17:00-17:15
An Evaluation On Securing Cloud Systems based on Cryptographic Key Algorithms
Sam Njuki, Jianbiao Zhang, Edna Too and Richard Rimiru
Beijing University of Technology, China
Abstract—The need for storage and movement of data in today’s ever increasing data
usage cannot be overemphasized. As a result, storage and movement of data within
the cloud becomes a viable option, even though this is a third party and some of the
data may be sensitive. Therefore, the security methods in place have to continually be
reviewed so as to get the best method or a combination of methods in this ever
changing and demanding field. In this paper, the authors have done a formal
assessment and examination of cloud systems key cryptographic algorithms and give a
recommendation of the best combination of algorithms as a result of our different
parameters of considerations. These parameters include whether they are a stream or
block cipher and their key size or the hash value. The authors found out that the most
viable cryptographic methods for securing the cloud systems are AES, Blowfish,
ABSTRACT
26 / 30
Diffie-Hellman, RSA and SHA-1 which is from symmetric, asymmetric and hashing
cryptographic algorithm respectively. They further noted that a hybrid of the three
cryptographic algorithms is more effective than an exclusive use of one cryptographic
algorithm. They propose a combination of AES, Diffie-Hellman and SHA-1 algorithm for
optimum cloud security.
AC037
17:15-17:30
Optimization of Data Distribution Strategy in Theta-join Process based on Spark
Shijiu Cao, Haihong E, Meina Song and Ken Zhang
Beijing University of Posts and Telecommunications, China
Abstract—The theta-join between tables is a common operation in the data query and
statistical analysis. When dealing with large amounts of data, it will produce a great
deal of cost. The theta-join inevitably generates huge computing and communication
overhead during data processing in the distributed environment. Besides, due to the
diversity of data, it also brings about the problem of data skew. In order to solve
uneven data distribution in theta-join and data skew in data processing, we propose a
solution, which can improve the data filtering strategy and put forward a data
distribution method using some affecting factors of data join efficiency quantified by
us. Our solution is implemented based on the distributed computing framework Spark.
The experimental results show that our method can be used for many types of data
and also shows better performance.
AC051
17:30-17:45
SLA-based Dynamic Virtual Machine Consolidation in Cloud Data Centers
Cheng-Ming Kuo, Han-Peng Jiang and Wei-Mei Chen
National Taiwan University of Science and Technology, Taiwan
Abstract—High energy consumption of data centers is an important issue which is
widely discussed. After dynamic virtual machine (VM) consolidation, the energy
consumption can be further reduced by switching the idle servers to sleep mode. Since
the resource demand of each VM varies over time, it may cause service level
agreements (SLA) violations between cloud service providers and users. In this paper,
we propose a dynamic VM consolidation algorithm to reduce the energy consumption
of data centers while meeting the quality of service (QoS) requirements. More
precisely, the proposed algorithm reduces both the number of migrations and the
energy consumption while keeping low and stabilized SLA violations. Compared to the
state-of-the-art algorithm, the proposed algorithm has a great improvement in the
number of migrations, SLA violations, and execution time.
AC207
17:45-18:00
An Application of Bilevel Optimization in Pricing and Leasing Strategy
Suthee Thitiananpakorn and Pisit Jarumaneeroj
Chulalongkorn University Bangkok, Thailand
Abstract—This paper focuses on the finding of an optimal pricing and leasing strategy
for both the third-party warehouse and customers, whose conflicts lie on their
contradictory objective functions, via bilevel optimization approach. Bilevel
ABSTRACT
27 / 30
optimization is a game-based approach where the third-party warehouse is a leader,
who first decides the price of short and long-term rental contracts, taking the
behavioral reaction of other players, that is, customers and its competitor, into
consideration. Once observed, customers then decide which contracts to be chosen so
that their objectives are optimized. While complicated, under certain assumptions, we
can show that a closed-form solution could be derived. Our experimental results
indicate that competitor storage price significantly affects both third-party
warehouse’s profit and all of its customer’s total costs. Besides, large customer tends
to suffer more from high opportunity cost; but, the delivery cost charged by the
third-party warehouse is crucial for small customer’s rental decision.
AC046
18:00-18:15
Correlation between Entrance Exam Scores (Stanine) and Academic Performance
Maria Cristina M. Ramos
Lyceum of the Philippines University – Batangas, Philippines
Abstract—Freshmen applicants are given an entrance examination to determine
readiness for tertiary education in all fields. The result of the test is a determining
factor whether the applicant is accepted or not in the school or program from which
the applicant seeks admission. Schools may opt to use a standardized test which may
be purchased from an outside vendor or from a teacher-made test which can be
validated within by the officer in-charge. Academic achievement represents
performance outcomes that indicate the extent to which a person has accomplished
specific goals that were the focus of activities in instructional environments, specifically
in school, college, and university.
The study used the correlation research design to determine the extent to which two
factors are related, not the extent to which one factor causes changes in another
factor. Respondents are first year to third year students of school year 2015-2016. A
total of 69 students who took the entrance examination and enrolled under the
Bachelor of Science in Computer Science (BSCS) program served as the respondents of
the study. The study used several variables such as the entrance examination results
(Stanine) and students’ academic performance (final grade) in 25 BSCS professional
courses.
AC049
18:15-18:30
Bayesian Networks in Intelligent Tutoring Systems as an Assessment of Student
Performance using Student Modelling
Roselie Alday
Lyceum of the Philippines University – Batangas, Philippines
Abstract—This paper developed a model for designing an intelligent tutoring system
for any programming language using Bayesian networks. The design model of the
tutoring system considers a user model using student model. The Bayesian network
was used to assess the current state of knowledge of the student so that the model can
adjust and present new knowledge to improve student performance as an outcome in
an e-learning environment.
ABSTRACT
28 / 30
Poster Session
AC006
Architecture modeling of new C2 System of joint anti-ship combat
Zhang Zhuang, LI Lin-lin and YU Hong-feng
Research Inst. of High-Tech, China
Abstract—By analyzing the department of defense architecture framework (DoDAF),
the architecture modeling process for C2 System of Joint anti-ship combat (C2JAC)
based on DoDAF2.0 is proposed. Combining with 0perational requirements of joint
anti-ship combat and specific operational process, key model of operational and system
perspective for C2JAC is designed by using the method of combining structured and
object-oriented. Finally, the behavior logic of the architecture dynamic model is verified
by the Rational Rhapsody architecture verification tool. The results express the
rationality of the designed architecture model.
AC016
Multi-core processor scheduling algorithm under the influence of process variation
Xiaohui Wei, Lishuang Su and Jingweijia Tan
Jilin University, China
Abstract—Due to process variation, different cores of multi-core processors will
generate frequency heterogeneity, and the difference in frequency will cause difference
in the speed of executing instructions. The existing multi-core processor scheduling
algorithms do not take the process variation into account, and ignore the deviation in
the manufacturing process. This paper proposes a frequency-heterogeneous scheduling
algorithm MFF (Max-Fast-First), which allocates long time-consuming tasks to cores
with high frequency and assigns short time-consuming tasks to cores with low
frequency, thus shortening the total operation time. The experimental results show
that the MFF algorithm reduces the running time by 17.87% compared to the dynamic
Min-Min algorithm, and it reduces the running time by 9.10% compared to the
CATS(Criticality-Aware Task Scheduler) algorithm. At the same time, the running time of
the MFF algorithm is less affected by the process variation, which is better than
dynamic Min-Min and CATS to process variation.
AC205
A Label Propagation Algorithm Based on Local Density of Data Points
Xiao Zhang, Zhixin Ma, Guoliang Deng, Qijuan Sun, Jingyi Zhang and Jun Yan
Lanzhou University, China
Abstract—Cluster analysis is one of the hot issues in the field of data mining and it has
extensive applications in many aspects. The label propagation algorithm is easy to
implement. At the same time, it has a low time complexity which has been recognized
by scholars. Because the algorithm needs to specify the category labels of the data set,
the accuracy and adaptability of the algorithm are affected. In view of the above
ABSTRACT
29 / 30
problems, this paper proposes a new clustering algorithm that combines the
advantages of density-based and label propagation. The algorithm adaptively
determines the label of the data points through local density and reducing the effect of
noise on the results. Experimental results show that the proposed algorithm has better
adaptability while improving the accuracy of clustering results.
AC050
High-Resolution Remote Sensing Scene Classification Using Improved LBP and SDSAE
Feng’an Zhao, Xiaodong Mu, Zhaoxiang Yi, Zhou Yang and Shuyang Wang
Xi'an Research Institute of High-Tech, China
Abstract—Classifying high-resolution remote sensing scene images with high accuracy
land-use is the challenging issues, the key of scene classification is to find effective
features of the scene image. The low-level visual feature methods, such as local binary
pattern (LBP), assume the same type of scene should share certain statistically holistic
attributes and have demonstrated their efficiency on scene classification. In this paper,
we propose an effective improved LBP method, called IRELBP, which use radial
difference and angular difference as its difference based descriptors to represent HRRS
scene images. Due to the fact that low-level features cannot represent more
meaningful semantic information, we extract features from a Stack Denoising Sparse
Autoencoder (SDSAE) to obtain more meaningful hierarchical feature. Both the global
features and hierarchical features are encoded by Fisher Vector, and then they are
concatenated into a discriminative representation. Finally are fed into the SVM
classifier for training or testing. We perform comprehensive experiments on two
remote sensing scene classification benchmarks: UC-Merced dataset and the recently
introduced large scale aerial image dataset (AID). The result demonstrates that our
proposed combination method provides an effective and discriminate feature
representation and outperforms the state-of-the-art methods in HRRS scene image
classification.
AC204
Parallel Optimization of Relion: Performance Comparison based on Cluster for
CPU/GPU and KNL
Heng Zhou, Fuchuan Ni, Liang Zhao, Fang Zheng
Huazhong Agricultural University, Wuhan, China
Abstract—Relion is the 3D reconstruction program with the Bayes algorithm of
electron cryo-microscope (cryo-EM) data. We analyzed the characteristics of the Relion
program, and designed a parallelization scheme. We use the optimization methods
commonly used in the code optimization for Relion programs, such as memory access
optimization, multithread optimization, vectorization and conversing coarse grained
parallel to fine-grained parallelism. Finally, the overall running time of the entire Relion
program was reduced by 379s. At the same time, we tested the program on GPU and
KNL platform and compared the results of the Relion program on the KNL cluster
platform and the GPU cluster platform. The results show that the optimization effect of
Relion on the GPU platform is better than KNL.
ABSTRACT
30 / 30
FP019
A Semi-Blind Receiver for Multi-Hop MIMO Relay Systems
Zhiqiang Yu, Minghu Gao and Yantao Lan
Yunnan Minzu University, China
Abstract—In this paper, we propose a novel semi-blind receiver for multi-hop
multi-input multi-output (MIMO) relay systems using the parallel factor (PARAFAC)
model. At the source node, A Khatri-Rao space-time (KRST) coding scheme is
considered to construct the transmitted signal. Each relay amplifies the received signals
and forwards the amplified signals to the next node. The received signals at the
destination node can be formulated as a PARAFAC model, and then a semi-blind
receiver is proposed for joint symbol and channel estimation by using an alternating
least squares (ALS) algorithm. Under the assumption that channel state information
(CSI) is not available at any node, our semi-blind receiver can implement joint channel
estimation and symbol decoding. Simulation results demonstrate the effectiveness of
the proposed semi-blind receiver.
Dinner <18:45-20:30> Location: Restaurant (2nd floor)
Note: dinner coupon is needed for entering the restaurant.